Chapter 1153 A Gift from Heaven
After listening to Chang Haonan’s interlocking schemes, the expression on Ding Zhongli’s face was very exciting for a moment.
If we sum it up in the way of later Internet era, it would be...
6.
Not only borrowing a knife to kill someone, but also borrowing someone else’s own knife.
The key is to cleverly exploit the psychological blind spots of most people. Before the final and most fatal output is played, it is almost impossible for anyone to prepare in advance.
However, it is not completely without problems.
Because the premise of all this is that a convincing enough algorithm can be produced so that the whole world, at least the academic community, can accept the methane concentration results derived from spectral results.
Otherwise, you will end up shooting yourself in the foot.
However, the integrity of academia on climate and environmental issues...
It can only be said that it is better than the public opinion circle.
But not much.
In fact, even the two most famous comprehensive academic journals in the world, Science and Nature, have clear positions on many issues.
If the submitted research results are inconsistent with the position behind the publisher, then unless it is a god-level paper that is irrefutable, it will most likely be rejected.
Therefore, Ding Zhongli still seemed a little worried:
"Academician Chang, of course I believe in your theoretical abilities... But climate research is not the same as engineering projects. The cycle of observing the evolution of the earth is generally based on 10 years, and it is difficult to come up with a sufficiently effective experiment through experimental means. The result of persuasion..."
"What I worry about is that the other party will directly encourage the academic community to oppose the inversion algorithm itself, and the media and the public are obviously incapable of judging whether the algorithm is correct or not. In the end It will still become the case that whoever makes the louder voice is justified... You know, we generally don't have any advantage in this regard..."
His statement is still too conservative.
China does not have an advantage in terms of public opinion.
It was like being killed to the point of losing one's armor.
However, Chang Haonan did not expect to rely on domestic media from the beginning:
“If a new atmospheric composition inversion method is directly proposed, the purpose is indeed too obvious... Even if the other party cannot directly object, They will also be vigilant in advance, so that we lose the suddenness of the attack.”
“So I plan to start from the theoretical level that is not directly related to climate issues. After all, this is my old profession. It still has some influence... After the mathematical and engineering circles have basically recognized the theoretical cornerstone of the inversion algorithm, more specific algorithms that are more targeted and suitable for the field of atmospheric remote sensing will be introduced. "
See. Seeing the other party's confident look, Ding Zhongli immediately understood.
This should have been drafted a long time ago.
He originally wanted to ask what the specific theory was, but after careful consideration, he felt that this was not his area of expertise. It would be a bit embarrassing if the other party explained a lot of results and he didn't understand them all.
So I simply skipped the middle part:
"Then after I go back this time, I will sort out the research results accumulated in the past in advance and prepare to assist you, Academician Chang?"< br>
“You can’t say that...”
Chang Haonan quickly waved his hand:
"Director Ding has been working hard on the climate issue and arguing for more than ten years. If he becomes a auxiliary just because I intervened once, wouldn't it appear that I am taking the lead?"
In this matter, , he was not prepared to stand directly on the bright side.
First of all, it is true that the division of labor between the two parties is different, and the IPCC meetings are basically held in Europe, and it is impossible for Chang Haonan to go abroad casually.
More importantly, the AE1500 in a few years will need to take low emissions as one of its core selling points.
So he has to be a red face now so that he can talk about environmental issues later.
Otherwise, if others investigate, they will find that your kid was the one who interfered with the IPCC assessment report and the climate agreement in the past few years, and the boomerang will hit their own heads...
In short, two Everyone was very satisfied with the meeting.
After some more mutual compliments, they left separately...
...
After returning to the laboratory, Chang Haonan immediately turned on the computer and clicked on the home page of the document. Wrote a series of titles——
"An optimized Levenberg-Marquardt iterative algorithm for optimizing initial guesses and global convergence of damping coefficients. 》
This is the first step in his entire plan.
Also the most important step.
For satellite remote sensing, forward modeling is equivalent to describing the physical process between atmospheric parameters and observed values, that is, using the physical process of radiation transfer to simulate the radiance of the sensor. Correspondingly, the inversion process is to establish a cost function under the optimization theory and gradually approach the true value in an iterative form.
Specific to the problem of atmospheric composition detection, in addition to the parameters that can be directly obtained by the sensor, the cost function also includes the error covariance matrix representing the observed radiation, the initial value of the unknown target vector, and the covariance matrix that can accurately describe the prior value of the background field. .
Because the second term can use the "known" carbon dioxide concentration as a reference, it does not completely require empirical guessing. Therefore, the essence of the inversion problem becomes to seek the minimum value of the cost function.
The most effective method that Chang Haonan can think of in a short period of time, but which does not arouse special vigilance, is to improve the Levenberg-Marquardt iterative algorithm at a theoretical level.
After all, this algorithm can play a role in many iteratively solved optimization problems.
This also includes machining and fluid dynamics calculations.
As a mathematician with a background in computing science, it is normal for him to study this kind of subject.
In fact, if there is more preparation time, Chang Haonan would be willing to consider some fast algorithms based on neural networks.
Unfortunately, neural networks have relatively high requirements for spectral resolution and require a large amount of statistical data for model training.
Considering that the IPCC’s fourth assessment will begin at the end of the year at the latest, we chose to seek stability.
In fact, under normal circumstances, the normal order of writing a paper is the main body first, then the conclusion, then the abstract, and finally the title.
However, Chang Haonan already had a rough idea of the entire content of the article, so after deciding on the title, he wrote the summary directly in order.
But after writing only a few sentences, he stopped typing on the keyboard——
If he were the only author of this article, even if he was the Philippine Prize winner, The title seems a bit thin.
Moreover, when Ding Zhongli uses it as a theoretical basis, others can also argue that "this is all a Chinese conspiracy" and so on.
But if we can bring some foreign institutions into the water...
Thinking of this, he immediately picked up the phone receiver at the corner of the table and called Liang Shaoxiu who was in Haojing...
...
"So...you mean, Can the system point (ERS) thermal deformation error during aircraft assembly be optimized?”
A few days later, Didier Casore looked at a research report written in two languages. , showing a somewhat surprised look.
He has already experienced China’s theoretical foundation in aircraft pre-assembly.
But even in the production plan at that time, the ERS point, as an enhanced reference system point for calibration, still needed to be manually calibrated by workers.
It has only been more than half a year, and it seems a bit outrageous that even this step can be done without manual labor.
However, his surprise did not last long.
“Of course it’s not possible now.”
Liang Shaoxiu shook his head:
"But on a theoretical level, this is indeed a possibility..."
Casore took the bait, showing a somewhat struggling expression——
After the success of the pre-production at the beginning of the year, People in the company ask every day whether similar technology can be used on Falcon 8Z or Rafale.
Even the Airbus Group next door has people coming to make insinuations.
This feeling of enjoying the stars and the moon is still very good.
Unfortunately, he could only give a negative answer every time, and eventually let the other party leave with disappointed eyes.
But now, there is an opportunity to follow up the research from scratch...
It is a joy that falls from the sky.
It’s hard not to be moved.
However, Kasore still had some self-awareness after all, and did not respond directly, but put on a distressed expression:
"But I'm an engineer, not a mathematician..."
"Neither am I!"
Liang Shaoxiu interrupted the other party pretending to be excited:
"So I It is suggested that in the name of our two companies, we jointly sponsor a team composed of Chinese and French researchers and let them solve theoretical problems... In this way, we can focus on the later production and testing parts and save a lot of time. Less resources and time..."
(End of this chapter)